Tài liệu Trilemma and macro economic policy choice in vietnam

1 MINISTRY OF EDUCATION AND TRAINING UNIVERSITY OF ECONOMICS HO CHI MINH CITY --------------- Đinh Thi Thu Hong TRILEMMA AND MACRO ECONOMIC POLICY CHOICE IN VIETNAM Major: Finance - Banking Code : 62340201 SUMMARY OF DOCTORAL THESIS Supervisor: Prof. Dr. Tran Ngoc Tho Ho Chi Minh City - 2014 2 Introduction 1. Necessity of the thesis The impossible trinity theory states that a country simultaneously may choose at most two of the following three goals: monetary independence, exchange rare stability and fully financial integration. In the context of financial integration steadily increasing in most countries, the choices and trade-offs between policy objectives become more and more important, because a combined policy will bring different effects for the economy. Therefore, the identification of the policy effectiveness will be very important for government agencies and macro administration to develop and implement appropriate policies in order to achieve certain economic goals. Thus, I selected research topic “Trilemma and macroeconomic policy choice in Vietnam” for my doctorate thesis. 2. Prior researches Many both domestic and foreign researches has exploited the impossible trinity theory issues in many different aspects such as the measurement methods of policy objective acheivement, impact of combined policies on to macroeconomic variables. However, studies on the impact of policies generally focused on inflation and output growth, not on unemployment rate (as Aizenman et al. 2008, 2010, 2011); or researches on Vietnam (such as some articles of Pham Thi Tuyet Trinh (2010), Le Phan Thi Dieu Thao (2010), Nguyen Tran Thuc Anh (2010), Nguyen Dai Lai (2013) …) mainly interpreted the the trilemma situation, have not analyzed the role of the policy combination on the economy, and not quantify the impacts, so obviously not much policy contribution provided. Differently, this thesis aimed at calculating the three index of trilemma for ten developing countries in Asia, especially Vietnam. The impacts of the policy combination on output growth, inflation and unemployment variables were also studied in this thesis. The empirical results of the thesis may imply some useful policy recommendations for the government of Vietnam. 3 3. Objectives of the thesis Objectives of the thesis are expressed through the following research questions: - How were the macroeconomic policy choices of Vietnam and nine other Asian countries from the perspective of the economic “trilemma” hypothesis, from 2000 to 2012? - Do the policy choices of countries in the sample have to be constrained by the impossible trinity theory or not? - How did the policy choices and the foreign reserves affect output growth volatility, the volatility of inflation, and unemployment rate, with focus on the countries in sample? - Do the level of financial development and the government expenditure affect the relationship between policies and economic stability or not? 4. Data The sample of the thesis consisted of 10 Asian countries because of the limitations of the data for the case of Vietnam while econometric models often require a relatively large number observations.The sample constructed on the basis of similarities among countries for drawing common as well as particular implications for Vietnam. The observation period was from 2000 to 2012. 5. Methodology In this thesis, I applied the methodology of Aizenman et al. (2008), Ito and Kawai (2012) to calculate the de factor trilemma indexes for Asean countries in the sample. The research models were mainly based on Aizenman et al. (2010). However, my study differed from them by testing the impact of policy choices on unemployment rate, along with output growth and inflation variables. In addition, I also examined the role of two moderator variables - financial development level and government 4 expenditure - on policy combinations. Different estimation methods with panel data (on Stata 11 sofware) were applied in the thesis appropriately. 6. The thesis contributions Different from previous studies on the same subject, the thesis has some new contributions: - Based on new measures of Aizenman, Chinn and Ito (2008), Ito and Kawai (2012), the thesis estimated the trilemma indexes: monetary independence (MI), exchange rate stability (ER) and financial openness (FO) for Vietnam and nine other Asian countries for the period of 2000 2012. The de factor measurements were suitable for testing the trilemma trade-offs. And this study finds that the policy choices in the ten countries in the sample were still binding as Mundell – Flemming Theory. - In order to examine the policy choices for open economies including Vietnam and nine other Asian countries, I used a sample similar to some of other prior studies but the research purpose is different. The thesis aimed at analyzing the impact of policy choice to macroeconomic performance (i.e output and inflation volatility, average of output, average of inflation, and average of unemployment rate) which has not ever been implemented before. - This is the first study investigating how each or any pair combination of the three policy choices affect the average of unemployment rate and significant results for Vietnam individually as well as the whole sample were achieved. The thesis is also the first one to fully calculate the threshold of international reserve (IR) holding to achieve specific economic goals for policy selection in practice. - The regression results of some control variables also contributed many notable findings, such as the impact of trade openness, TOT shock on output volatility, the impact of 2008 financial crisis on unemployment rate, and the role of external financing. - Not only examining the interaction between the trilemma choices and financial development, and their influences on the economic stability, but the thesis also finds that government expenditure affects the link between trilemma policy configurations and output volatility, inflation volatility. 5 - Based on empirical findings of the trilemma theory together with Vietnam's economic situation, some policy choices are recommended for Vietnam in order to maximize the benefits and to reduce risk from financial integration, to stabilize and grow the economy. 7. Structure of the thesis Chapter 1: Theoretical framework and empirical evidence Chapter 2: Data and Methodology Chapter 3: Empirical results and discussion Chapter 4: Policy reccommendation for Vietnam Chapter 1 Theoretical framework and empirical evidence 1.1. From IS – LM to Mundell-Fleming model: 1.1.1. Effectiveness of fiscal and monetary policy under fixed exchange rate The Mundell–Fleming model, was also known as the IS-LM-BP model, describing effectiveness of fiscal and monetary policy under fixed exchange rate. Accordingly, under the fixed exchange rate regime and free capital flows, fiscal policy is highly effective while the monetary policy is ineffective. 1.1.2. Effectiveness of fiscal and monetary policy under floating exchange rate In a floating exchange rate regime and free capital flows, the monetary policy has a greater effect while fiscal policy has little effect. 1.2. The impossible trinity theory 6 The impossible trinity theory states that a country simultaneously may choose at most two of the following three goals: monetary policy independence, exchange rare stability and fully financial integration. 1.3. The extensive studies of impossible trinity theory 1.3.1. Development of the trilemma dimensions and the diamond patterns Foreign reserves is considered as the fourth peak in triangular illustrating the impossible trinity theory. The earlier literatures focused on the role of international reserves as a buffer stock for exchange rate stability while maintaining pretty high level of monetary policy independece and financial liberalization. 1.3.2. The methodology for the construction of “trilemma indexes”  Monetaty policy independence  Exchange rate stability  Financial market openness 1.3.3. The role of international reserves Foreign exchange reserves plays an important role in helping countries achieve higher levels of the trilemma policies. In other words, international reserves can mitigate the policy trade-off of trilemma policies. It also plays an important role in protecting financial markets from capital flows reversal and keeping the exchange rate in the target band. 1.3.4. Trilemma configurations of countries in sample Many erlier studies analyzing the policy choices in the countries and regions through different periods confirmed that impossible trinity theory had become an important indication for policy makers. The value of the theory remains unabated even when many countries had to reconsider and change their policy choices after the crisises, or in current unstable economic and political conditions. 7 1.3.5. The relationship between policy choices and macroeconomic variables Monetary policy, exchang rate regime and financial liberalization policy are always important tools with that, a government could achieve its growth and economic stability objectives. Many empirical studies have examimed the effects of policy choices on the macroeconomic variables. In order to add in to prior researches’ contribution, and to provide empirical evidences for the case of Vietnam and other nine Asian countries, the thesis especially analyzed the impacts of combined policies to unemployment (along with output growth and inflation variable). In addition, I calculated the de factor “trilemma indexes” and tested the validity of a linear specification of these indexes to examine whether the notion of the trilemma can be considered to be a trade-off and binding. Chapter 2 Data and methodology The sample of this thesis included ten Asean countries with some comparable characteristics1. They are China, Hong Kong, India, Indonesia, Korea, Malaysia, Phillipines, Singapore, Thailand and Vietnam in the period of 2000 – 2012. Hong Kong, Korea and Singapore are high income countries, – as defined by the World Bank. The data set is primarily from the World Bank and IMF sources, and processed by three years rolling window. 2.1. Regression on panel data 2.1.1. The basic estimations Pool regression model Fixed effect model Random effect model 1 The choice of countries in the sample was similar to Patnaik, et al. (2011), Patnaik and Shah (2010), except for Taiwan because of lack of data. 8 Test for appropriate model After selecting the most appropriate model from three basic models, the thesis tested hypotheses about the phenomenon of heterokedasticity and autocorrelation. Test for heterokedasticity and autocorrelation Because the phenomenon of heterokedasticity and autocorrelation simultaneously existed in the model, the thesis used FGLS regression on panel data to fix them. 2.1.2. FGLS regression 2.1.3. Regression with dummy variables 2.2. Data and methodology 2.2.1. Examining the linear relationship between the three trilemma indexes Linear regression model: 1 = β1MIi,t + β2ERi,t + β3FOi,t + t (2.1) where: MI is the monetary independence index, ER is the exchange rate stability index, FO is the financial integration index; i refers to countries and t represents the time index. 2.2.2. Examining the impact of policy choices to the macroeconomic variables Basic model: Yit = α0 + α1ITit + α2IRit + α3(ITit x IRit) + α4FCi + EFitB + XitГ + Ziф + ɛit (2.2) where: Yit: vector of dependent variables, is the measure of macro policy performance for country i in year t. ITit (Impossible Trinity): a vector of any two of the three trilemma indexes, namely MI, ER, and FO. 9 IRit (International Reserve): the level of international reserves (excluding gold) as a ratio to GDP. ITit x IRit : an interaction term between the trilemma indexes and the level of international reserves. FCi (Financial Crisis): a dummy for financial crisis. EFit (External Financing): vector of external financing sources to country i in year t. Xit : a vector of macroeconomic control variables (relative income, trade openness, TOT shock… ). Zt : a vector of global shocks. ɛit : error term. i : country index t : time index (yearly, from 2000 to 2012) 2.2.3. Description of variables Dependent variables Independent variables Other control variables Table 2.1: Expectation of correlations in the model Order Number Variables 01 MI 02 ER 03 FO 04 Reserve (IR) 05 Crisis 06 FDI 07 FPI 08 Others capital 09 Short-term debt Output growth volatility Average output growth Inflation volatility Average Inflation Average unemployment rate + +/+ + + +/+ + +/+/+ +/+/+ + + + + + + + +/+ + + + + +/+/+ + +/+/- 10 10 TDS 11 Relative income 12 Trade openness 13 TOT shock 14 Fiscal policy 15 Volatility of M2 16 Private credit 17 Average of inflation 18 Inflation Volatility 19 US rate 20 World ouputgap 21 Oil shock + +/+/+ +/+/+/- - + +/+ - + +/+/+ + + +/+/+/+ +/- + +/+/- + + + +/+ 2.2.4. Analysis procedure The thesis began with calculating MI, ER, FO indices, based on methods of Aizenman, et al. (2008), Ito and Kawai (2012). After that, I estimated the linear regression (model 2.1) to test whether the three trilemma policy goals were linearly correlated, and confirmed the notion that a rise in one trilemma variable should be traded-off with a drop of a linear weighted sum of the other two. Next, I examine how the policy choices among the three trilemma policies (monetary independence, exchange rate stability, financial integration) affected output growth volatility, average of output growth, volatility of inflation, average of inflation, and average of unemployment, based on regression model 2.2. For each dependent variable, I considered three models of combined policies2. Each model was estimated with both the full sample of ten Asean countries and a subgroup of seven developing countries (exclude three high income countries: Korea, Hong Kong and Singapore). In order to test the different effects in the case of Vietnam, I used the Vietnam dummy variable (VN) in models with full sample, and compared the results. 2 Model 1: combination of MI and ER; Model 2: combination of MI and FO; Model 3: combination of ER and FO. 11 2.2.5. The method of calculating the impossible trinity indexes: Monetary policy independent index (MI) Exchange rate stability index (ER) Financial integration index (FO) 2.3. Descriptive analysis 2.3.1. The policy choices of Asean countries in the sample:  The degree of monetary policy independence (MI) of countries in the sample Figure 2.2 illustrates the degree of monetary policy independence (MI) of countries in the sample  The degree of exchange rate stability (ER) of countries in the sample Figure 2.5 illustrates the degree of exchange rate stability (ER) of countries in the sample  The degree of financial integration (FO) of countries in the sample Figure 2.7 illustrates the degree of financial integration (FO) of countries in the sample 2.3.2. The policy combinations in the context of trilemma: Figure 2.8 illustrates the impossible trinity configuration in ten Asean countries 2.3.3. The level of foreign exchange reserves and diamond patterns: Figure 2.11 shows diamond patterns of the Asean countries in the sample Chaper 3 Empirical results and discussion 12 3.1. Empirical results of testing the linearity among MI, ER, FO The results of regression model (2.1) are extremely good, reflected in the very high adjusted R-squared figures (95% for full sample, and 97% for subsample of developing countries). These results imply a linear correlation among the impssible trinity indexes, and countries in the sample actually had to face with the trilemma trade-offs. All coefficients are significant at 1%. 3.2. Empirical results of testing the impacts of policy choices to macroperformance 3.2.1. Impacts on the volatility output growth   The regression results for the full sample and the subsample. The regression results for the full sample with dummy variable for Vietnam. Table 3.4: Summary of the impacts on the volatility of output growth Model 1 Model 2 0.10 0.05 - MI ER (0.05) (0.10) 10 7 nước nước Independent variables MI ER Model 3 0.20 0.15 0.10 0.05 (0.05) (0.10) (0.15) 0.06 0.04 MI VN +/ unless IR>32% 10 nước 7 nước +/ unless IR>30% FO IR>46% 32%30% VN Model 3 Full Developing Vietnam Sample countries +/ unless +/ unless IR>43,8% IR>32% +/ unless IR>29% - Volatility of (0.02) Volatility of output growth Model 3 Model 2 Full Developing Vietnam Full Developing Vietnam Sample countries Sample countries -/ unless -/ unless -/ unless -/ unless IR>42% IR>100% IR>43% IR>34% +/ unless IR>46% To reduce the ER 10 7 nước nước VN FO 0.02 FO IR<100% IR<43% 29%43,8% % output growth Source: summarized and calculated from the Tables 3.2 and 3.3. +/ unless IR>50% +/ unless IR>24% - IR>32% IR>50% 13 Table 3.4 illustrates the impacts of trilemma indexes to ouput growth volatility, for each combination of the two policy choices. Monetary independence indicator (MI) generally has a negative effect on output growth volatility; greater exchange rate stability (ER) is likely to induce output growth volatility as expectation. On the other hand, greater financial integration (FO) comes at the cost of higher output growth volatility. These effects can change when combined with the level of international reserves (IR). 3.2.2. Impacts on the average of output growth  The regression results for the full sample and subsample  The regression results for the full sample with dummy variable for Vietnam. Table 3.7: Summary of the impacts on the average of output growth 0.15 Model 1 0.10 0.15 0.05 0.10 - MI (0.05) ER (0.10) 10 7 nước nước MI ER VN Model 3 0.20 0.15 0.05 MI - FO (0.05) 0.10 FO 0.05 ER - (0.10) (0.15) Independent variables Model 2 0.20 (0.15) (0.05) 10 nước 7 nước 10 nước 7 nước VN Average of output growth Model 2 Model 2 Full Developing Vietnam Full Developing Vietnam Sample countries Sample countries -/ unless -/ unless -/ unless IR>34% IR>30% IR>24% +/ unless IR>56% +/ unless IR>56% +/ unless IR>48% FO To increase the34%30% - +/ unless -/ unless IR>39% IR>67,5% 24%67,5% Average of output growth Source: summarized and calculated from the Tables 3.5 and 3.6. VN +/ unless IR>39% + - +/ unless IR>23% IR<39% IR<23% 14 Table 3.7 illustrates the impacts of trilemma indexes to average ouput growth, for each combination of the two policy choices. Monetary independence indicator (MI) generally has a negative effect on average output growth; greater exchange rate stability (ER) is likely to increase average output growth. Greater financial integration (FO) is mostly to lower average output growth, except for Vietnam interaction term. These effects can change when combined with the level of international reserves (IR). 3.2.3. Impacts on the volatility of inflation The regression results for the fullsample and the subsample The regression results for the fullsample with dummy variable for Vietnam.   Table 3.10: Summary of the impacts on the volatility of inflation 2.00 Model 1 1.00 0.50 (0.50) MI (1.00) ER (2.00) (4.00) Volatility of - FO (0.50) ER (1.00) 10 nước 7 nước VN Model 1 Full Developing Vietnam sample countries -/ unless -/ unless IR>50% IR>108% -/ unless -/ unless IR>43,5% IR>42% Volatility of inflation Model 2 Full Developing Vietnam sample countries -/ unless IR>121% -/ unless IR>33,5% IR<43,5% IR<42% 10 7 nước nước VN +/ unless IR>25% FO To reduce the FO (1.50) 10 nước 7 nước ER 0.50 MI (2.00) (3.00) Model 3 1.00 (1.00) (1.50) MI 1.50 - - Independent variables Model 2 1.00 25%33% 33%25% -/ unless IR>48% -/ unless IR>29% +/ unless IR>32% IR<25% 32%13% IR<13% 15 Table 3.10 illustrates the impacts of trilemma indexes on the volatility of inflation, for each combination of the two policy choices. Monetary independence (MI) and exchange rate stability (ER) mostly have a negative effect on the volatility of inflation; while financial integration (FO) has different effect in each sample. These effects can change when combined with the level of international reserves (IR). 3.2.4. Impacts on the average of inflation   The regression results for the fullsample and the subsample The regression results for the fullsample with dummy variable for Vietnam Table 3.13: Summary of the impacts on the average of inflation 0.40 Model 1 0.40 Model 2 0.20 0.30 0.50 MI 0.20 ER 0.10 MI (0.20) FO (0.40) 10 nước 7 nước Independent variables MI ER 10 7 nước nước VN Model 1 Full Developing Vietnam sample countries +/ unless + + IR>58% +/ unless +/ unless IR>54% IR>34,8% IR>54% IR>34,8% FO (1.00) ER VN (2.00) 10 nước 7 nước Average of inflation Model 2 Full Developing Vietnam sample countries - + FO To reduce the (0.50) (1.50) (0.60) - IR>58% Model 3 1.00 -/ unless -/ unless IR>26% IR>32% IR<26% IR<32% + Full sample VN Model 3 Developing Vietnam countries +/ unless IR>26% +/ unless IR>23% +/ unless IR>18% -/ unless -/ unless -/ unless IR>18% IR>25% IR>29% 23%89% IR>42% ER -/ unless IR>42% -/ unless IR>41% -/ unless IR>40% FO To reduce the -/ unless IR>38% -/ unless IR>34% + + +/ unless IR>58% -/ unless IR>69% IR<34% IR<69% +/ unless -/ unless IR>65% IR>20,6% IR<41% IR<40% IR>65% Model 3 Developing Vietnam countries IR<38% Average of unemployment rate Source: summarized and calculated from the Tables 3.14 and 3.15. 17 Table 3.16 illustrates the impacts of trilemma indexes to the average of unemployment rate, for each combination of the two policy choices. Monetary independence indicator (MI) has a positive effect on the average of unemployment rate; exchange rate stability (ER) has negative effect, except for Vietnam interaction term in model 3. Greater financial integration (FO) contributes to increase unemployment rate in subgroup of developing Asean countries, except for Vietnam. These effects can change when combined with the level of international reserves (IR). 3.3. Interactions between the Trilemma configurations and Financial development, Government expenditure Because the fact that importance of two control variables (private credit and fiscal policy) in the basic model (2.2) has been proven but seemed insignigicantin my thesis, I further analyzed the role of Financial development and Government expenditure to the relationship between the trilemma configurations and macroeconomic stability. On the basis of the model (2.2), combined with expectations of the role of financial development, government spending to macroeconomic stability, I adjusted the model as follows: Yit = γ0 + γ1PCit + γ2(ITit x IRit) + γ3(ITit x PCit) + uit Yit = λ0 + λ1GEit + λ2(ITit x IRit) + λ3(ITit x GEit) + νit (3.1) (3.2) where: PCit (Private Credit) – PC high/ medium/ low: representing the level of financial development, is a dummy variable of private credit creation as a ratio to GDP of country i in year t, for different level groups of PC high/ medium/ low. GEit (Government Expenditure) – GE high/ medium/ low: a dummy variable of government expenditure as a ratio of general government final consumption expenditure to GDP of country t in year t, for different level groups of GE high/ medium/ low. 18 3.3.1. The role of Financial development   Effects of trilemma indexes on Ouput volatility Effects of trilemma indexes on Inflation volatility 3.3.2. The role of Government expenditure   Effects of trilemma indexes on Ouput volatility Effects of trilemma indexes on Inflation volatility 3.3.3. Financial development and government expenditure in Vietnam Policy choices of Vietnam from 2000 to 2012: the thesis results showed that Vietnam has selected independent monetary policy combined with a stable exchange rate for the period from 2000-2003, and from 2008-2010; the remaining years in sample, it was in favor of a stable exchange rate policy combined with financial liberalization. The degree of financial development in Vietnam from 2000 to 2012: based on the ratio of private credit/ GDP and classification as described in Section 3.3.1, Vietnam was ranked as medium financial market development country (excluding 2009 and 2010 at high level). The empirical analysis results imply that if Vietnamese government aims at reducing the volatility of real income per capita, the policy combination of exchange rate stability and monetary independence would be the choice under the condition of a medium financial market development. Otherwise, in order to stabilize inflation, the choice of combining independent monetary policy and exchange rate stability proves more feasible. The degree of government expenditure in Vietnam from 2000 to 2012: According to the calculations of the thesis based on data from the Worldbank and classifications described in Section 3.3.2, Vietnam was ranked as the low government expenditure country. The empirical analysis results imply that if Vietnamese government aims at reducing the volatility of real income per capita, the policy combination of monetary independence and financial liberalization would be the choice 19 in the context of low government expenditure. Otherwise, in order to stabilize inflation, the choice of exchange rate stability could be followed. Chapter 4 Policy reccomendation for Vietnam 4.1. Applying the impossible trinity theory in the policy selection Based on the empirical results of the thesis, the development of financial liberalization and the data of current foreign reserves of Vietnam (about 32 billion US dollars at the end of 2013, equivalent to about 20% of GDP), I suggest policy combinations for each specific economic objectives: - Stabilizing growth of real income per capita: greater monetary independence and a higher financial liberalization (Model 2) - based on the results in Table 3.3 and 3.4; - Increasing real income per capita: exchange rate stability and enhancing financial liberalization (Model 3) - based on the results in Table 3.6 and 3.7; - Stabilizing inflation rate: greater monetary independence and a higher financial liberalization (Model 2) - based on the results in Table 3.9 and 3.10; - Reducing the average inflation: greater monetary independence and a higher financial liberalization (Model 2) - based on the results in Table 3.12 and 3.13; - Reducing the average of unemployment rate: exchange rate stability and enhancing financial liberalization (Model 3) - based on the results in Table 3.15 and 3.16; These results indicate that none of policy combinations is perfect to help the Vietnamese government simultaneously achieve all macroeconomic objectives. However, to achieve most of the goals, the model 2 combination of monetary independence and financial integration - is better over all (although this option did not consider the requirements of foreign exchange reserves). This choice is also appropriate with the inevitable 20 trend of increasing financial liberalization in developing countries, not only in Vietnam. So that, the government should reduce its priorities for exchange rate stability and create more space for monetary independence and financial integration. Based on the conditions of economic indicators in Vietnam, compared to other countries in the sample (Figure 4.1), the priority order of objectives should be: stabilize output growth, stabilize inflation rate, lower inflation rate, improve income and lower unemployment rate. 4.2. Enhance the reputation and independence of the State Bank  Independence of policy objectives  Increasing transparency and accountability  Independence of the central bank personnel policy 4.3. Increasing the flexibility of the exchange rate In a fixed exchange rate regime, that serves as a nominal anchor for monetary policy, the monetary policy will lose their independence and can not control inflation or deal with disturbances along with large capital inflows. Therefore, in the context of increasing financial integration, more flexible exchange rate should be the appropriate choice for a more monetary independence policy. 4.4. Controling risks of financial integration  Maintain control of capital flow  Develop a healthy domestic financial markets  Strengthen market confidence 4.5. Enhance the accumulation of international reserves